arkindex/process/api.py

@@ -478,7 +478,7 @@ class FilesProcess(CreateAPIView):
         folder_type = serializer.validated_data.get('folder_type')
         element_type = serializer.validated_data['element_type']
 
-        if folder and folder.corpus != corpus:
+        if folder and folder.corpus_id != corpus.id:
             # The files' corpus is already validated as writable
             raise ValidationError({'__all__': ['Element and files are in different corpora']})
 

arkindex/process/serializers/imports.py

@@ -24,7 +24,6 @@ from arkindex.project.validators import MaxValueValidator
 from arkindex.training.models import ModelVersionState
 from arkindex.users.models import Role
 from arkindex.users.utils import get_max_level
-from transkribus import TranskribusAPI
 
 
 class ProcessLightSerializer(serializers.ModelSerializer):
@@ -242,7 +241,10 @@ class ProcessListSerializer(ProcessLightSerializer):
 class FilesProcessSerializer(serializers.Serializer):
 
     mode = EnumField(ProcessMode, default=ProcessMode.Files)
-    files = serializers.PrimaryKeyRelatedField(queryset=DataFile.objects.all(), many=True)
+    files = serializers.PrimaryKeyRelatedField(
+        queryset=DataFile.objects.select_related('corpus'),
+        many=True,
+    )
     folder_id = serializers.UUIDField(required=False, allow_null=True)
     folder_type = serializers.SlugField(required=False, allow_null=True)
     element_type = serializers.SlugField()
@@ -253,7 +255,7 @@ class FilesProcessSerializer(serializers.Serializer):
         'unique_corpus': 'Imports can only run on files from a single corpus',
         'corpus_read_only': 'Cannot write in corpus',
         'folder_not_found': 'Folder does not exist',
-        'image_or_pdf': 'File imports can only import images or PDF documents',
+        'unsupported_content_type': 'File imports can only import images, PDF documents or ZIP archives',
         'iiif_only': 'IIIF imports can only import IIIF documents',
         'folder_required': 'Either folder_type, folder_id or both are required',
         'iiif_folder_required': 'IIIF imports require both folder_type and element_type',
@@ -274,7 +276,7 @@ class FilesProcessSerializer(serializers.Serializer):
         if len(corpora) > 1:
             self.fail('unique_corpus')
         corpus = corpora.pop()
-        if corpus not in Corpus.objects.writable(self.context['request'].user):
+        if not Corpus.objects.writable(self.context['request'].user).filter(id=corpus.id).exists():
             self.fail('corpus_read_only')
         return files
 
@@ -291,8 +293,12 @@ class FilesProcessSerializer(serializers.Serializer):
 
     def validate(self, data):
         if data['mode'] == ProcessMode.Files:
-            if not all(f.content_type == 'application/pdf' or f.content_type.startswith('image/') for f in data['files']):
-                self.fail('image_or_pdf')
+            if not all(
+                f.content_type in ('application/pdf', 'application/zip', 'application/x-zip-compressed')
+                or f.content_type.startswith('image/')
+                for f in data['files']
+            ):
+                self.fail('unsupported_content_type')
 
         elif data['mode'] == ProcessMode.IIIF:
             if not set(f.content_type.split(';')[0] for f in data['files']) <= {'application/json', 'application/ld+json'}:
@@ -621,52 +627,6 @@ class CorpusProcessSerializer(serializers.Serializer):
         return data
 
 
-class ImportTranskribusSerializer(serializers.Serializer):
-    """
-    Serialize a Transkribus import
-    """
-    collection_id = serializers.IntegerField(min_value=1)
-
-    def validate(self, data):
-        collection_id = data.get('collection_id')
-
-        # Check Transkribus email
-        transkribus_email = self.context['request'].user.transkribus_email
-        if not transkribus_email:
-            raise serializers.ValidationError(
-                {"__all__": ["You have not registered your transkribus email"]}
-            )
-
-        # Login as Arkindex user on transkribus
-        try:
-            transkribus_client = TranskribusAPI(email=settings.TRANSKRIBUS_EMAIL, password=settings.TRANSKRIBUS_PASSWORD)
-        except Exception:
-            raise serializers.ValidationError(
-                {"__all__" : [f"Failed to login on Transkribus as {settings.TRANSKRIBUS_EMAIL}"]}
-            )
-
-        # Check Arkindex's right
-        try:
-            users = transkribus_client.list_user_collection(collection_id)
-        except Exception:
-            raise serializers.ValidationError(
-                {"collection_id" : [f"User {settings.TRANSKRIBUS_EMAIL} is not a member of the collection {collection_id}"]}
-            )
-
-        # Check user's right
-        user = next(filter(lambda user: user["email"] == transkribus_email, users), None)
-        if not user:
-            raise serializers.ValidationError(
-                {"collection_id": [f"User {transkribus_email} is not a member of the collection {collection_id}"]}
-            )
-
-        return data
-
-
-class CreateImportTranskribusErrorResponseSerializer(serializers.Serializer):
-    collection_id = serializers.CharField(required=False, help_text="Errors that occurred during collection ID field validation.")
-
-
 class ProcessElementLightSerializer(serializers.ModelSerializer):
     """
     Serialises an Element, using optimized query for ListProcessElement

arkindex/process/tests/test_processes.py

@@ -58,6 +58,16 @@ class TestProcesses(FixtureAPITestCase):
             size=42,
             content_type='application/json',
         )
+        cls.zip_df = cls.corpus.files.create(
+            name='test.zip',
+            size=1337,
+            content_type='application/zip',
+        )
+        cls.windows_zip_df = cls.corpus.files.create(
+            name='windows.zip',
+            size=1337,
+            content_type='application/x-zip-compressed',
+        )
         cls.page_type = ElementType.objects.get(corpus=cls.corpus, slug='page')
         cls.volume_type = ElementType.objects.get(corpus=cls.corpus, slug='volume')
         cls.ml_class = cls.corpus.ml_classes.create(name='clafoutis')
@@ -1902,20 +1912,35 @@ class TestProcesses(FixtureAPITestCase):
         self.assertIsNone(process.element)
 
     @override_settings(IMPORTS_WORKER_VERSION=None)
-    def test_from_files_image_and_pdf(self):
+    def test_from_files_multiple_types(self):
         self.client.force_login(self.user)
-        with self.settings(IMPORTS_WORKER_VERSION=str(self.import_worker_version.id)):
+
+        with self.settings(IMPORTS_WORKER_VERSION=str(self.import_worker_version.id)), self.assertNumQueries(30):
             response = self.client.post(reverse('api:files-process'), {
-                'files': [str(self.pdf_df.id), str(self.img_df.id)],
+                'files': [
+                    str(self.pdf_df.id),
+                    str(self.img_df.id),
+                    str(self.zip_df.id),
+                    str(self.windows_zip_df.id),
+                ],
                 'mode': 'files',
                 'folder_type': 'volume',
                 'element_type': 'page',
             }, format='json')
             self.assertEqual(response.status_code, status.HTTP_201_CREATED)
+
         data = response.json()
         process = Process.objects.get(id=data['id'])
         self.assertEqual(process.mode, ProcessMode.Files)
-        self.assertListEqual(list(process.files.all()), [self.img_df, self.pdf_df])
+        self.assertQuerysetEqual(
+            process.files.all(), [
+                self.img_df,
+                self.pdf_df,
+                self.zip_df,
+                self.windows_zip_df,
+            ],
+            ordered=False,
+        )
         self.assertIsNone(process.element)
 
     @override_settings(
@@ -1925,7 +1950,7 @@ class TestProcesses(FixtureAPITestCase):
     def test_from_files_iiif(self):
         self.client.force_login(self.user)
 
-        with self.assertNumQueries(28), self.settings(IMPORTS_WORKER_VERSION=str(self.import_worker_version.id)):
+        with self.assertNumQueries(27), self.settings(IMPORTS_WORKER_VERSION=str(self.import_worker_version.id)):
             response = self.client.post(reverse('api:files-process'), {
                 'files': [str(self.iiif_df.id)],
                 'mode': 'iiif',
@@ -2033,14 +2058,17 @@ class TestProcesses(FixtureAPITestCase):
 
     def test_from_files_files_wrong_type(self):
         self.client.force_login(self.user)
-        response = self.client.post(reverse('api:files-process'), {
-            'files': [str(self.iiif_df.id)],
-            'folder_type': 'volume',
-            'element_type': 'page',
-            'mode': 'files',
-        }, format='json')
-        self.assertEqual(response.status_code, status.HTTP_400_BAD_REQUEST)
-        self.assertEqual(response.json(), {'non_field_errors': ['File imports can only import images or PDF documents']})
+
+        with self.assertNumQueries(6):
+            response = self.client.post(reverse('api:files-process'), {
+                'files': [str(self.iiif_df.id)],
+                'folder_type': 'volume',
+                'element_type': 'page',
+                'mode': 'files',
+            }, format='json')
+            self.assertEqual(response.status_code, status.HTTP_400_BAD_REQUEST)
+
+        self.assertEqual(response.json(), {'non_field_errors': ['File imports can only import images, PDF documents or ZIP archives']})
 
     @override_settings(IMPORTS_WORKER_VERSION=None)
     def test_from_files_folder_id(self):